Category Technology to Design Intelligent Systems for Complicated Decisions

A new technology is proposed to design AI systems using authors’ theory of category approach to development, description and research of AI systems in the frames of a rather general and universal setting. The corresponding fundamental base is exposed in the present paper with the use of Lisp computer language allowing direct implementation demonstrated in several examples. The bases of the proposed technology rely on a universal representation of the production as a pair of patterns combined with some concretization operation, which is responsible for an applicability of considered productions to the data that were made available in the intelligent system during its performance. The main purpose of the created technology is providing the user with a possibility to construct a complicated intelligent system on the level of thinking over its general concept, yet obtaining at the end the performing AI system with a guaranty of implementation of all the requirements that have been formulated by the system creator in the process of its development, or follows from the proposed technology. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.